Comprehensive Untargeted Metabolomic Analysis from Five Stages of Blueberry Development by High Resolution Accurate Mass Spectrometry

Significance of the topic

Understanding metabolite changes during fruit development is essential for optimizing crop quality, nutritional value, and processing performance. Comprehensive untargeted metabolomics reveals biochemical shifts that simple targeted workflows can miss, informing cultivar selection, harvest timing, and post-harvest handling. This study examines how combining complementary chromatographic separations and MS acquisition strategies expands metabolome coverage across five developmental stages of blueberry (Vaccinium corymbosum).

Objectives and study overview

• Compare metabolomic profiles across five blueberry developmental stages to identify stage-specific metabolite populations.
• Evaluate the effect of chromatographic modes (RP and HILIC) and MS acquisition strategies (DDA and DIA, positive and negative ESI) on feature detection and library identifications.
• Recommend optimal method combinations to maximize coverage while considering analysis time.

Methods

Blueberry samples were collected from four-year-old Vaccinium corymbosum plants grown at Texas A&M University. Samples representing five defined developmental stages were prepared and extracted following published protocols. Analyses were performed in randomized triplicates, with pooled quality-control (QC) samples prepared by combining aliquots from all individual extracts to monitor system performance and reproducibility.

Used instrumentation

• Liquid chromatography: Shimadzu Nexera HPLC (Nexera X3).
• Mass spectrometry: Shimadzu LCMS-9050 Q-TOF operated with electrospray ionization (ESI) in both positive and negative polarities.
• Chromatography modes: reversed-phase (RP) and hydrophilic interaction liquid chromatography (HILIC).
• MS acquisition modes: data-dependent acquisition (DDA) and data-independent acquisition (DIA), resulting in eight analytical workflows (RP/HILIC × positive/negative × DDA/DIA).
• Data processing: LabSolutions Insight Profiler for feature alignment, component detection, PCA, and F-statistic ranking; library matching against MS-DIAL, EMBL, and GNPS libraries.

Main results and discussion

• System performance and reproducibility: Overlayed MS1 total ion chromatograms from pooled QCs demonstrated excellent chromatographic and MS reproducibility for both HILIC and RP methods, confirming platform stability across batches.
• Feature detection: Positive ESI mode yielded more aligned features and detected components than negative mode. HILIC separations produced a larger number of features and components compared with RP, indicating strong complementarity between the techniques.
• Acquisition mode comparison: DIA and DDA produced comparable numbers of features and components, but combining both chromatographic and fragmentation strategies improved overall identification coverage.
• Library matching: Matches were performed against three libraries; more than 75% of library identifications were captured by combining HILIC DDA and RP DDA in both polarities, suggesting an efficient reduced set of methods for broad coverage.
• Multivariate patterns: PCA across all eight analytical workflows consistently separated sample groups, with early developmental Stage 1 showing the greatest differentiation from later stages. This indicates major metabolic transitions occur early in fruit development.
• Representative metabolites: Specific examples included thiamine, which showed substantially higher variability in Stage 1 versus later stages, and sayaendoside, which varied significantly across multiple stages. Box plots and mirrored MS2 spectra supported confident library matches for selected compounds.

Benefits and practical applications of the method

• Enhanced metabolome coverage: Combining HILIC and RP separations with both polarities and complementary MS acquisition strategies uncovers metabolites otherwise missed by single-mode workflows.
• Stage-specific biochemical insight: Distinct metabolite populations across developmental stages can inform optimized harvest timing to maximize desired bioactives (e.g., antioxidants, vitamins).
• Efficient screening strategy: For laboratories prioritizing throughput, pairing HILIC DDA and RP DDA captures the majority of library-identifiable metabolites while reducing run complexity and time.
• Quality control and reproducibility: Pooled QC monitoring and randomized analysis order support robust comparative studies and reduce batch-related artifacts.

Future trends and potential applications

• Integration with ion mobility: Adding ion mobility separation would improve isomer discrimination and reduce spectral congestion, increasing confident identifications.
• Expanded and curated libraries: Growth of MS/MS spectral repositories and improved in silico fragmentation tools will raise identification rates for plant metabolites.
• Targeted follow-up: High-confidence library hits from untargeted screening can be validated and quantified in targeted assays for biomarker development or quality control metrics.
• Multi-omics and spatial metabolomics: Combining metabolomics with transcriptomics/proteomics and spatially resolved techniques could map metabolic pathways and tissue-specific dynamics during fruit maturation.
• Standardization and automation: Harmonized sample prep, standardized acquisition templates, and automated data workflows will enable wider adoption in breeding, agronomy, and food science labs.

Conclusion

The study demonstrates that a complementary analytical strategy—pairing RP and HILIC chromatographies with both positive/negative ESI and multiple MS acquisition modes—substantially improves untargeted metabolome coverage across blueberry developmental stages. All five stages exhibited distinct metabolic signatures, with early development showing the largest differences. For laboratories balancing coverage and throughput, combining HILIC DDA and RP DDA provides an effective compromise that captures the majority of library-identifiable metabolites. The approach supports discovery-driven metabolomics to guide agricultural and nutritional decisions, while further gains are expected from enhanced libraries, orthogonal separations, and integrated multi-omics workflows.

References

1. Das PR, Darwish AG, et al. Diversity in blueberry genotypes and developmental stages enables discrepancy in the bioactive compounds, metabolites, and cytotoxicity. Food Chemistry. 2022; 374:131632. doi.org/10.1016/j.foodchem.2021.131632.